Resting-State EEG Microstates Across a Dimensional Spectrum of Autistic Traits: From Typical Development to Diagnosed ASD

Autism spectrum disorder (ASD) has been linked to atypical large-scale brain dynamics, but it is unclear how these alterations extend across the broader autism phenotype. We applied a seven-class resting-state EEG microstate model (A–G) to adults with clinical ASD and to typically developing adults with high (TD-High) and low (TD-Low) autistic traits, quantified with the Autism-Spectrum Quotient. We compared temporal parameters, spatial coverage, explained variance, and both observed and chance-corrected transition probabilities. Across all microstates, the ASD group showed a globally more fragmented regime than both TD groups, with markedly shorter but more frequent microstate episodes and reduced duration variability. By contrast, TD-High and TD-Low were similar on these global indices. At the network level, Microstate C showed reduced explained variance and coverage in ASD relative to both TD groups. In Microstates E and G, explained variance and coverage increased from TD-Low to TD-High to ASD, with TD-High consistently occupying an intermediate position. Mean GFP and GFP variability for Microstate E were also elevated in ASD relative to both TD groups. Transition analyses revealed reduced short-range transitions within an early A–C ensemble and increased transitions from these states into other microstates in ASD, with TD-High again showing an attenuated, intermediate pattern. Chance-corrected transitions confirmed that sensory/self-related routes occurred less often than expected, whereas routes from these states into other microstates were over-expressed. These findings support a dimensional account in which EEG microstates index autism-related network organisation across clinical and subclinical ranges.